Ljubljana/Strategy

Strategy

The HIV virus mutates at a rate of 3x10^-5 per nucleotide per replication cycle and with generation of 10⁹ to 10¹⁰ virions every day it is clear that viral mutability has to be bypassed in order to develop an efficient antiviral therapy.

Requirements for the effective antiviral treatment:

• should be BASED ON VIRAL FUNCTION and INSENSITIVE TO MUTATIONS
• should be VERSATILE - allow activation of different effectors, e.g.

  - those that kill infected cells that spread infection (activation of apoptosis)

  - destroy virus (RNase, APOBEC3...)

  - activate human immune defense system (e.g. interferons, chemokines…)

• system should be able to respond to low levels of activation – the RESPONSE SHOULD BE AMPLIFIED
• avoid activation in noninfected cells

Considered approaches

• use of siRNA - attractive but not suitable because of the mutations
• antibodies (same problem)
• use of viral LTR promoter - experimental evidence that transcription is "leaky"
• use the chimeras with TLR3 intracellular domain, which activates the interferon beta signaling pathway - problem with active homodimers (tested, high constitutive activity)
• use viral reverse transcriptase to integrate the effector into the genome (difficult, perhaps next time :-) )
• use of viral protease activity (OK, works)
• detect formation of CD4-CCR5/CXCR4 heterodimer (OK, works)

Components of the antiviral defense device

Detection / Activation
We decided to concentrate on two viral functions - binding to cellular receptors and processing of viral proteins by HIV protease.
For detection of viral binding to the cells we took advantage of the formation of the CD4 and CCR5 (or CXCR4) coreceptor dimer upon binding of viral gp120 protein. We looked for ways to detect the formation of heterodimeric membrane proteins and couple this event to the cellular response. We found that the split ubiquitin system has been previously used in yeast to detect the interaction between membrane proteins (Stagljar and Fields, 2002), which is exactly what we needed, except that its function has not yet been described for mammalian membrane proteins. To be safe, the other, parallel approach was to use the split TEV (tobacco etch virus) protease system (Wehr et al., 2006), which restores the proteolytic activity upon heterodimerization of their fusion partners. HIV protease activation could be detected by specific cleavage of a protein containing a linker with a protease substrate site similar to the one previously reported before by introduction of a substrate protein into HIV-infected cells (Vocero-Akbani et al, 1999).

Amplification
Detection of infection of a cell with a single viral particle should provoke a significant response, therefore the signal would have to be amplified. This can be achieved by biological processes which employ different enzymes (e.g. activation of a prodrug by reconstituted lactamase...) or through coupling to cellular transcription/translation machinery by different RNA polymerases. We have also considered membrane anchoring of DNAse linked to a nuclear localization signal. After cleavage off the membrane, as a result of activation, DNase would migrate into the nucleus and trigger apoptosis. We opted for the T7 RNA polymerase as it specifically binds to the specific promoter sequence and is commonly used in biotechnological applications for bacterial overexpression of proteins. Additionally its functionality in mammalian cells has also been described (Lieber et al, 1989). The response of this amplification mechanism can even be enhanced by the addition of an autoamplifiable BioBrick with positive autofeedback, which thus increases the amount of available T7 RNA polymerase. The switch to turn on the T7 RNA polymerase was achieved by anchoring it to the membrane and adding the NLS signal to direct it to the nucleus after the activation step. The activation step is reached after receptor dimerization, which results in proteolytic degradation of the linker between the membrane anchor and T7 polymerase.

Activation of defense effector
What type of response do we want the cells to have upon infection by a virus?
One possibility is to simply kill the cell, thereby preventing the virus from forming hundreds of copies and spreading the infection. Other possibilities include activation of an organism's defense system using e.g. interferons, or enzymes that destroy viral components, such as APOBEC3, RNase and others. Activation of the defense system by caspases can only be used in leak-proof systems and is probably most effective when it is already too late for other types of defense. It is a great advantage if the same platform can be used to deploy different types of effectors -- this is exactly what systems controlled by T7 promotor provide. Similarly, transcription factors from other cell types (e.g. yeast) could be used as well, as long as they provide the specific responsiveness to activation, resulting from the viral detection.